Happy Thursday and welcome.

GP evaluation has come up in several conversations recently.

Not necessarily the detailed underwriting of a fund. More the question that comes before it:

How do you decide which managers deserve your time?

It is an evergreen problem for LPs. There are always more managers worth meeting than there are hours in the calendar. And because the consequences of missing a great manager can be significant, the natural response is to keep taking meetings.

That works for a while.

It becomes harder once you have met 500 managers, receive hundreds of updates each year, and still need time to diligence the small number of funds you may actually invest in.

At that point, “meet as many managers as possible” is no longer a strategy.

You need a system.

The one I used across my career was simple: every manager received a letter from A to E.

The letter was not the investment decision. It was a decision about attention.

Note: The below is an updated version of this original post, with new views, articles and context added.

I have previously written about the framework I used to evaluate managers across five areas:

Each area was scored from one to five.

That analysis helped us understand the manager. But it did not answer a more practical question:

What should we do next?

The A-E rating sat on top of the underlying analysis and translated it into an action.

We assigned the rating after the first meeting and reconsidered it after every meaningful interaction.

An A-rated manager was one we expected to commit to unless the remaining work revealed a major surprise.

These managers had:

We met them at least once a year (or tried to).

The purpose was to maintain the relationship, follow developments, and ensure we were ready when the next fund came to market. And, of course, to quietly confirm our rating.

A B-rated manager was genuinely in contention, but one or more important questions remained.

I have previously written about our Alpine miss, and I am fairly certain we rated Alpine a B after our initial introductions.

These managers stayed active in the pipeline, but the cadence was lower. We might meet every 12–18 months and follow developments through quarterly letters, annual meetings, references, and market conversations.

A B could become an A.

But that required new evidence - not simply another fundraising meeting covering the same material.

C was probably the most useful category.

These were credible managers that did not currently look competitive for our portfolio.

Perhaps the performance was average. Perhaps the fund was too large. Perhaps we liked the strategy but could not identify a meaningful edge.

Nothing was necessarily wrong.

But “nothing is wrong” is not the same as “we should spend time here.”

C-rated managers remained in the system. We might meet them every 18–24 months or when something material changed.

What we did not do was continue taking meetings simply because the GP asked.

That distinction saved a lot of time.

A D rating meant we had identified several unresolved concerns and were unlikely to invest in the current fund.

We would usually remain polite and keep the relationship open. Teams change. Strategies evolve. Performance develops.

But the manager no longer received regular meeting time.

We would follow asynchronously and reconsider the rating if something genuinely changed.

The word “genuinely” matters here.

A new deck, a higher mark, or another quarter of the same narrative is not necessarily new evidence.

An E-rated manager was outside our mandate or a clear no.

No regular meetings. No active follow-up. No elaborate process.

Archive it and move on.

This sounds obvious, but many LP pipelines contain opportunities that everyone knows will never receive a commitment.

They remain active because nobody has explicitly closed them.

The value of the system was not that it made us better at assigning letters.

The value was that the letters changed what we did.

When planning a trip, we started with the A-rated managers in that city.

When a GP began fundraising, we could immediately see whether we had been following them closely enough to form a view.

When an internal discussion became vague, the rating forced us to be more precise.

Why was this manager a B rather than an A?

Why were we taking another meeting with a C?

Had anything materially changed since the last meeting?

The system also made it easier to say no without closing the relationship permanently.

We could explain that we had recently met, that our current view had not changed, and that we would prefer to reconnect after a specific milestone.

That is more honest than accepting another meeting with no realistic path to an investment.

When a B became an investment

In one case, we met a GP raising a small annex fund.

We liked the strategy and the team, but needed more evidence. We rated the manager B and recorded exactly what was missing.

Nine months later, we followed up. Some of the evidence had arrived, and the next fund opened shortly afterwards.

Because the manager had remained active in our system, we were ready to engage early.

When we kept chasing an A

In another case, we rated a manager A after the initial meeting.

Then we heard almost nothing for three years.

The system told us to keep showing up. Whenever we visited the GP’s city, we reached out. We sent occasional updates and maintained the relationship without creating unnecessary noise.

Eventually, an allocation became available.

The GP remembered that we had remained interested and consistent, even when there was nothing immediate for us.

We received the call and made the commitment.

Neither outcome came from a clever piece of analysis.

They came from following a process for longer than would have felt natural without one.

The rating only created value because it lived in the CRM and influenced the workflow.

At a minimum, we tracked four things:

None of these fields are complicated.

Together, they tell you which relationships matter, when something is likely to happen, when you last spoke, and who is responsible for the next step.

That is enough to turn a list of managers into a sourcing process.

A letter grade can create false precision.

Two B-rated managers may look completely different. One may be close to becoming an A. The other may be drifting toward C.

The rating should therefore never replace the underlying analysis or the written rationale.

It is a workflow tool.

Its purpose is to turn judgment into a consistent allocation of time.

The best managers are not always available when you want them to be. Relationships need to be built before a fundraise. Evidence needs to be collected over several years.

A good sourcing system helps you decide where that effort is most likely to matter.

The A–E framework was ours.

It was simple, occasionally imperfect, and required discipline to maintain.

But it helped ensure that our calendar reflected our investment priorities—not merely the order in which meeting requests arrived.

FundFrame is growing. This week, we’re welcoming Alfred as a developer.

I could write the usual paragraph about how excited we are to have him - and how his joining will benefit our customers.

Both are true.

Instead, I’ve been experimenting with AI image models recently. It is part fun, part nostalgia for me. I grew up with MS-Dos and Windows 3.11. Creating images in this style takes me back (as an AI would put it).

Below is my latest creation with our updated team slide.

For those interested in the details: it was made with ChatGPT 5.6 Terra, using medium reasoning.

This newsletter is written by Steffen Risager, the founder of FundFrame, a platform for LPs to manage their private markets investments.

Before that, Steffen was CIO at Advantage Investment Partners, a Danish Fund-of-Funds.

Steffen has a decade of experience as an LP, and has made commitments totalling approx. $6bn across fund- and co-investments.